Do People Check the Smell of Their Hands After a Handshake?

After a handshake, people sometimes casually bring a hand toward their face. That movement may serve to check a smell. A 2015 study found that, after shaking hands with someone of the same sex, people spent more time bringing their right, shaking hand near their nose than after a greeting without a handshake.

The researchers did more than film behavior. They also investigated whether skin-derived compounds transfer during a handshake, whether nasal airflow increases when a hand is at the face, and whether changing the odor around the other person changes the behavior. Their aim was to test the hypothesis that handshakes provide an opportunity to obtain olfactory information from another person.

A brief handshake transfers compounds from the skin

First, an experimenter wearing nitrile gloves shook hands with 10 volunteers. Each person took part three times over three days. The researchers analyzed compounds on the gloves used for handshakes and on gloves worn in the same way without a handshake.

After handshakes, squalene, palmitic acid, and geranyl acetone were detected for all 10 volunteers, but not in the no-handshake controls. The researchers also tested a hand soap used in common by participants, but found none of these three compounds. The authors therefore considered them likely to originate from the skin.

This established that even the brief contact involved in a greeting can transfer compounds from one person’s skin to another. This analysis alone does not establish whether those three compounds actually served as signals to humans, or which compounds the nose received.

People bring their hands to their face even while doing nothing

The main behavioral experiment recruited 175 people and analyzed 153 after excluding 22 for departures from the procedure, such as using a mobile phone. Participants waited alone in a room. About three minutes later, an experimenter entered and delivered a scripted greeting lasting about 20 seconds. Eighty participants received a handshake and 73 did not, after which they waited alone again.

The study included combinations of male and female participants with male and female experimenters, with 20 experimenters taking part. Participants were filmed without being told the purpose of the observation. Afterward, they could choose whether to allow the footage to be used in research or have it destroyed.

Two raters recorded, separately for each hand, the time spent touching the face below the eyebrows and above the chin. Already during the observation period before the greeting, 85 of the 153 participants touched their face. Touches were often near or below the nose. People made these movements quite frequently even without a handshake.

After a same-sex handshake, checking the right hand increased

The analysis compared periods of about 80 seconds before and after the greeting: one minute plus the duration of each person’s greeting. First, the researchers calculated each person’s change by subtracting the pre-greeting value from the post-greeting value. They then subtracted the change in the no-handshake group with the same participant–experimenter sex combination. This was intended to account for baseline face-touching tendencies and changes associated with the greeting itself.

The results showed an increase in time spent bringing the right hand used in the handshake near the nose after a same-sex handshake. In a supplementary comparison that did not subtract the change in the no-handshake group, the mean before-to-after change was +2.14 seconds with a handshake and −5.39 seconds without one; the between-group difference was significant.

In opposite-sex greetings, a difference appeared for the left hand, which had not been used in the handshake. The before-to-after change in time spent bringing the left hand toward the face was +2.91 seconds with a handshake and −7.95 seconds without one. Here, the comparison was enlarged not only by an increase after handshakes but also by a decrease in the no-handshake group.

In other words, this was not a simple pattern in which people sniffed both hands equally regardless of whom they shook hands with. In this greeting setting, the movements of the two hands differed depending on the sex combination of the two people.

When a hand was near the nose, inhaled airflow also increased

Face-touching can also occur for reasons such as nervousness. In a separate group of 33 people, the researchers therefore measured airflow using a thin tube at the nose. Among the 17 who brought a hand toward their face during measurement, airflow was greater when the hand was near the nose than at baseline. The mean increased from about 113 to 238 mL per second, a statistically significant difference.

This supports the idea that at least some of the observed movements involved drawing in air to sample an odor, rather than simply touching the face. Wearing the apparatus reduced face-touching itself, so the airflow findings and the behavioral findings obtained without the apparatus need to be considered together as separate experiments.

Changing the odor around the other person also changed the movements

The researchers then restricted the experiment to handshakes between women and added odors around the female experimenter. They compared a group exposed to CK-be perfume and groups exposed separately to AND and EST, two steroid compounds studied as candidate chemical signals, with the earlier female–female group in which no odor had been added.

AND and EST were released into the air from a modified wristwatch. They were not applied directly to the hand used for shaking; the manipulation changed the odor around the experimenter.

The combination of changes in time spent bringing the left and right hands to the face differed by odor condition. With EST, for example, checking of the right hand decreased and checking of the left increased relative to the condition without added odor. The perfume group showed an increase for the right hand, but a significant perfume-versus-no-added-odor difference was not demonstrated.

The researchers had also expected that AND and EST might have opposite effects, but this prediction was not supported: the changes were in similar directions. Seven of 21 participants in the perfume condition and four of 20 in the EST condition reported noticing an odor. Most did not report noticing one, but the stimuli were not unnoticed by everyone.

Thinking of a handshake as a way to gather odor information

Combining the transfer of skin compounds, the increase in inhalation when a hand approached the face, and the changes in behavior following handshakes and odor manipulations, the authors concluded that handshaking may be a way of obtaining social olfactory information. They also proposed that people might check the other person’s odor with the right hand and their own with the left, making a comparison. That explanation still awaits testing.

The authors themselves also noted that the left–right differences between same-sex and opposite-sex encounters might be specific to this experimental setting. Familiarity, culture, and the purpose of the meeting were not investigated. Nor did the study measure the historical origins of handshaking or changes in liking or trust toward the other person.

Summary

In this study, handshakes transferred compounds thought to originate from the skin. After same-sex handshakes, people spent more time bringing their right hand near their nose than after greetings without a handshake. Separate airflow measurements showed increased inhalation accompanying the movement, and experiments adding odors also changed the pattern of left- and right-hand movements.

A small everyday gesture may provide an opportunity to check another person’s odor, or one’s own. This study explored that possibility through chemical analysis, behavior, and breathing.

The research was supported by the James S. McDonnell Foundation, Israeli Centers for Research Excellence, and the US Air Force Office of Scientific Research. The authors declared no competing interests.

Reference

Frumin, I., et al. (2015) A social chemosignaling function for human handshaking. eLife, 4, e05154. Original paper / Open full text